← Back to Portfolio
IDist ® — Asset Management Infrastructure
Lead Engineer & Architect Enterprise Infrastructure · Asset Management 2024–2025

IDist ®

IDist ® is a global design and procurement platform for physical asset management infrastructure — the system that tracks, validates, and coordinates the lifecycle of distributed physical assets across international operations. Built to replace fragmented spreadsheet-and-email workflows with a unified, auditable pipeline from asset specification through procurement, deployment, and verification.

The platform serves organizations managing large physical asset portfolios across multiple countries and regulatory environments, where procurement errors cascade into six-figure cost overruns and compliance failures compound across jurisdictions.

4 Countries Unified
30% Procurement Cycle Reduction
89% Fewer Specification Errors
100% Real-Time Asset Visibility
Featured in Google Startups Accelerator
Challenge Approach Architecture Impact Takeaways
01 — The Challenge

Asset management across borders, without a single system of record.

Managing physical asset portfolios across four countries means four regulatory frameworks, four procurement standards, four ways of tracking what exists and what doesn't — and no shared infrastructure connecting any of them. The client was running their global asset operations on a patchwork of spreadsheets, email chains, and disconnected legacy platforms that had accreted over years of regional expansion.

The consequences were predictable: procurement errors that weren't caught until assets arrived on-site, compliance documents that referenced specifications no longer in service, and no way to answer a basic question — "what is the current deployment status of this asset class across all our markets" — without a multi-day data gathering exercise.

4 Countries, 4 Systems
No shared data model, no unified identifier, no cross-market visibility. Every region managed assets independently.
$2.3M
Average annual procurement waste from specification mismatches — assets ordered, shipped, and rejected due to non-compliant specifications.
0%
Real-time visibility into asset deployment status. Leadership had no live view of what was deployed, in transit, or pending compliance review.
02 — Research & Approach

Normalize before you build. Then automate the hard parts.

The first and most important decision was to spend significant architecture time on the data model before writing application code. International asset management requires a canonical data model that normalizes specifications across different national standards — without that foundation, every automation layer built on top creates new inconsistencies.

The canonical model aligned asset specifications to ISO 55000 (the international asset management standard), with jurisdiction-specific extension fields for each market's regulatory requirements. This meant that a single asset specification could be used to generate the correct procurement documents for any market the client operated in.

  • Designed a unified asset data model aligned to ISO 55000, with per-jurisdiction extension schemas for regulatory fields
  • Built a vendor compliance scoring engine that validates suppliers against the specific requirements of each jurisdiction before allowing them into a procurement workflow
  • Implemented real-time procurement tracking with automated milestone verification — each stage of the procurement process has defined completion criteria that must be met before the workflow advances
  • Created a specification matching engine using structured comparison against validated asset catalogs, flagging mismatches before orders are placed rather than after assets arrive
  • Designed an automated document processing pipeline that ingests vendor submissions (invoices, certifications, spec sheets) and validates them against the expected format and content for each procurement stage
03 — Technical Architecture

GCP-native, event-driven, auditable by design.

The stack was chosen for auditability and maintainability at enterprise scale. Every state change, every approval, and every specification match is logged with full provenance — not as a compliance afterthought, but as a core architectural requirement that shaped the data model from the start.

Frontend
Next.js on Cloud Run with role-based access control — separate views for procurement teams, engineering teams, and compliance officers. Each role sees only the workflows and data relevant to their function.
API Layer
FastAPI on Cloud Run. All procurement state transitions go through the API, which enforces business rules and writes audit events before committing state changes.
Processing
Cloud Functions for event-driven validation — triggered on document upload, specification changes, and procurement state transitions. Document AI + Gemini API for vendor document extraction and normalization.
Data
BigQuery for the asset registry and analytics (immutable append-only audit log). Firestore for real-time procurement state (mutable, watched by frontend). Cloud Storage for document archive.
Integration
Connector layer for ERP systems, vendor portals, and compliance databases. Each integration is stateless and writes only to the canonical data model — no direct coupling between external systems.
GCP Cloud Run BigQuery Next.js Python FastAPI Document AI Gemini API Firestore Cloud Functions
04 — Measured Impact

Fewer errors, faster cycles, and a view of the whole operation.

30%
Procurement Cycle Reduction
End-to-end procurement time dropped from an average of 12+ weeks to under 9, driven by automated validation and milestone tracking eliminating manual back-and-forth at each stage.
4
Countries on One Platform
All markets migrated to the unified platform, with jurisdiction-specific compliance fields handled automatically. No regional workarounds or parallel systems.
89%
Reduction in Specification Errors
The specification matching engine catches mismatches before orders are placed. Prior to the platform, 23% of procurement orders contained specification errors that required rework.
Live
Real-Time Asset Visibility
Leadership now has a live view of every asset's status across all markets — deployed, in transit, in procurement, or pending compliance review — updated in real time as procurement events occur.
05 — Key Takeaways

What this project taught about building for enterprise operations.

"Normalize Before You Optimize"
International asset management requires a canonical data model before any automation adds value. We spent approximately 40% of architecture time on the data model — defining the canonical asset schema, the jurisdiction extension pattern, and the identifier strategy — before writing a single line of application code. That investment paid back at least 10x in reduced edge cases, reduced integration complexity, and reduced bug surface during implementation.
"Compliance Is a Feature, Not a Constraint"
Building jurisdiction-aware validation into the core pipeline — not as an afterthought layer added after the core system was built — eliminated the most expensive failure mode: non-compliant procurement that requires physical rework after assets are already in the field. When compliance is structural, it's impossible to skip. When it's a wrapper, it gets bypassed under pressure.
"Enterprise Trust Requires Audit Trails"
Every state change, every approval, and every specification match is logged with full provenance — who did what, when, and why. This wasn't a compliance checkbox. It was the specific feature that got procurement directors to actually migrate off spreadsheets. The first question every enterprise stakeholder asks is: "if something goes wrong, can I see exactly what happened and who was responsible?" The answer had to be yes before anyone was willing to rely on the system for decisions that affected real money.

View More Work

See the full portfolio — production AI systems across insurance, property intelligence, and voice biometrics.